RE: [EXTERNAL] Re: Extending a /64 (The most welcome comment)

["Manfredi (US), Albert E" <[e-mail]>]

-----Original Message-----
From: ipv6 <[e-mail]> On Behalf Of Tony Whyman

> Over the last few days, I have been happy to try and peel away the 
issues that lay behind our proposed IPv6 addressing plan and to use it 
as an opportunity to spread understanding of the ATN/IPS and the 
constraints under which we are working. However, there is one point that 
it seems to be too difficult for some to get their head around and that 
is that we are not starting with a "clean sheet of paper".

That is what I've come to understand, from your explanations. Which is why something that should be quite straightforward, nothing more than standard mobile IPv6, becomes difficult. It will require perhaps a new set of prefixes from IANA, as well as an overlay to make the scheme actually work, because it is not the way IP is set up. Neither IPv4, nor IPv6, work this way.

The way I would be proposing to continue, assuming they insist on using IPv6 (which is *not mandatory*, of course), is to make a small adjustment in the philosophy of where the airplane ID should go. Or say it another way, wink-wink, pretend we are extending the prefix, when in fact, we are pre-assigning some of the upper IID bits. (But don't tell them that.)

1. Use straight out of the box mobile IPv6. Even use the convention so many care about, that IIDs must be 64 bits wide. (Path of least resistance.)

2. The prefix of each address is a prefix that belongs to the home country of the airline. This is because the home address, in mobile IP, must be routable.

3. Each airplane is allocated at least a /56 (for the front cabin and machinery), or even better, a /48.

4. The top 40 or 48 bits of the IID are used as airplane ID.

5. Addresses for these machinery devices are either assigned statically, or they are assigned using DHCP cum the "client identifier" option. By doing this, you will not need a DNS inside the airplane.

6. Anyone wanting to initiate a session with anything inside the cabin would reach the home agent first, and then either the packets are relayed to the current c/o address of the airplane, by the home agent, or the source encapsulates the packets and sends directly to the c/o address, per standard mobile IP.

Always a problem when those in charge try to reinvent wheels. Unfortunately.

Bert